CN111542068B - A Cognitive Perception Optimization Method for Simulating Primary User Attacks in Cognitive Networks - Google Patents

Abstract

The invention discloses a collaborative sensing optimization method for simulating primary user attacks on a cognitive network. The detection channel and the reporting channel in the collaborative sensing process are considered jointly, and the optimal K value is found through the K-rank optimization criterion to minimize the global average of the system. Error probability, and compared with traditional AND criterion, OR criterion, and majority criterion. The results show that the proposed scheme of the present invention effectively reduces the global average error probability of the system, significantly improves the accuracy of spectrum sensing, and helps to solve the security problem of wireless spectrum sensing under simulated primary user attacks.

Description

A Cognitive Perception Optimization Method for Simulating Primary User Attacks in Cognitive Networks

technical field

The invention belongs to the technical field of wireless communication, and in particular relates to an optimization method for spectrum sensing used for spectrum shortage.

Background technique

The development of mobile communication technology provides people with more and more powerful and convenient means of communication, which is profoundly changing people's production and way of life. However, with the rapid increase in the number of mobile users worldwide, the rapid growth of Internet services and the widespread use of portable computer equipment, the demand for wireless spectrum resources in communication systems is also increasing. Under the circumstance of limited natural spectrum resources, many countries have already allocated allocable spectrum resources, leaving little spectrum for new services and new technologies, or even no spectrum resources for allocation.

Traditional technologies can improve spectrum utilization and transmission capacity to a certain extent, but even so, the problem of spectrum resource shortage has not been effectively solved. Under this situation, the existing static spectrum allocation scheme obviously cannot meet the rapidly growing demand of high-speed wireless communication services. Therefore, new technologies need to be developed to provide more available spectrum for new services. Cognitive Radio can selectively utilize the idle wireless spectrum resources of the main user, and is considered to be an effective technology to solve the current shortage of wireless spectrum. The core idea is that CR has the ability to learn and interact with the surrounding environment to perceive and utilize the available spectrum of the space and limit and reduce the occurrence of conflicts.

A Cognitive User (CU) determines whether a Primary User (PU) is occupying a licensed frequency band through spectrum sensing. If the detection result shows that the PU does not occupy the licensed frequency band, the frequency band is allocated to other users for use. When only a single cognitive user CU is used to sense whether the primary user PU occupies the licensed frequency band, multipath fading will make the detection result unsatisfactory, resulting in an error in the decision result.

SUMMARY OF THE INVENTION

The technical problem to be solved by the present invention is: in the process of cognitive radio sensing and utilizing the available spectrum in space, the judgment result may be wrong, and the sensing precision is not high.

In order to solve the above technical problems, the present invention provides a collaborative sensing optimization method for simulating primary user attacks on cognitive networks. The collaborative spectrum sensing optimization system includes a primary user (PU), a simulated primary user attacker (PUEA), M Cognitive User (CU) and a Data Fusion Center (FC), including the following steps:

In step A, the false alarm probability of the energy detection of a single cognitive user is used as a constraint condition to determine the judgment threshold of the energy detection of the cognitive user. probability;

Step B, derive the detection probability and missed detection probability formula of cognitive user energy detection, and obtain the global average error probability formula of the collaborative spectrum sensing optimization system, the detection probability exists for the main user and the detection result of the cognitive user is also: The probability that the main user exists, the missed detection probability is the probability that the main user exists and the detection result of the cognitive user does not exist;

Step C, using the K-rank optimization criterion to calculate the global average error probability.

The global average error probability of the cooperative spectrum sensing optimization system is the superposition of the detection channel error probability and the reporting channel error probability, where:

Detection channel error probability includes false alarm probability and missed detection probability;

The reported channel error probability includes the transmission error probability.

In step A, the following four situations correspond to the four channel states:

( _a ) S1 assumption: the primary user and PUEA do not exist, only the case of noise;

(b) S2 assumption: there are only primary users and noise _;

( _c ) S3 assumption: there are only PUEA users and noise;

( _d ) S4 assumption: the case where the primary user, the PUEA user, and the noise all exist.

The main user and cognitive user signals both obey the Gaussian distribution, and the noise signal is additive white Gaussian noise.

Assuming that the channel environment between all cognitive users and the primary user is the same, and the information received by the cognitive users through energy detection has the same signal-to-noise ratio, the above four channel states correspond respectively:

S ₁ ={F ₀ , H ₀ }

S ₂ ={F ₀ , H ₁ }

S ₃ ={F ₁ , H ₀ }

S ₄ ={F ₁ , H ₁ }#(1)

Among them, F ₀ indicates that PUEA does not exist, F ₁ indicates that PUEA exists, H ₀ indicates that the main user does not exist, and H ₁ indicates that the main user exists; if PUEA detects the existence of the main user, PUEA initiates a simulated main user attack with probability α. If PUEA does not detect the existence of the primary user, PUEA initiates a simulated primary user attack with probability β;

The false alarm probability P _fc represents the probability that the primary user does not exist but the detection result of the cognitive user shows that the primary user exists, that is, P _fc =P(D ₁ |H ₀ ), which can be obtained from the Bayesian formula:

Among them, D ₁ indicates that the main user exists in the result of cognitive user energy detection, D ₀ indicates that the main user does not exist in the result of cognitive user energy detection, and P(F ₀ |H ₀ ) indicates that the PUEA does not exist when the main user does not exist. The probability of existence, P(F ₁ |H ₀ ) represents the probability of the existence of PUEA when the primary user does not exist, and P(D ₁ |F ₀ , H ₀ ) does not exist as the primary user and PUEA, but the detection result of the cognitive user is the probability of the existence of the primary user, namely:

P(D ₁ |F ₁ ,H ₀ ) represents the probability that the primary user does not exist, the PUEA exists but the cognitive user detection result is the primary user, that is:

t为积分变量，λ_c为认知用户能量检测的判决阈值；γ_e为PUEA在认知用户处的接收信噪比；

表示背景噪声的方差；N为采样点数；其中，P(F₀|H₀)、P(F₁|H₀)的表达式分别为：in,

t is the integral variable, λ _c is the decision threshold of cognitive user energy detection; γ _e is the received signal-to-noise ratio of PUEA at the cognitive user;

represents the variance of background noise; N is the number of sampling points; among them, the expressions of P(F ₀ |H ₀ ) and P(F ₁ |H ₀ ) are:

P(F ₀ |H ₀ )=P _fΔ ·(1-α)+(1-P _fΔ )·(1-β) (5)

P(F ₁ |H ₀ )=P _fΔ ·α+(1-P _fΔ )·β (6)

为PUEA对主用户进行检测时的虚警概率：

False alarm probability when detecting primary users for PUEA:

λ _Δ is the decision threshold for the PUEA user to perform energy detection on the primary user.

In step B, substitute formulas (5) and (6) into formula (2) to obtain:

表示主用户存在、且认知用户的检测结果也显示主用户存在的概率，P(F₀|H₁)表示主用户存在时PUEA不存在的概率，P(F₁|H₁)表示主用户存在时PUEA也存在的概率，即

由贝叶斯公式可得：Similarly, the detection probability

represents the existence of the primary user, and the detection result of the cognitive user also shows the probability of the presence of the primary user, P(F ₀ |H ₁ ) represents the probability that the PUEA does not exist when the primary user exists, and P(F ₁ |H ₁ ) represents the primary user The probability that PUEA also exists when it exists, i.e.

It can be obtained from Bayesian formula:

P(D ₁ |F ₀ , H ₁ ) represents the probability that the primary user exists, the PUEA does not exist, and the cognitive user detection result exists as the primary user, namely:

P(D ₁ |F ₁ , H ₁ ) represents the probability that the primary user exists, the PUEA also exists, and the cognitive user detection result exists as the primary user, namely:

The expressions of P(F ₀ |H ₁ ) and P(F ₁ |H ₁ ) are:

P(F ₀ |H ₁ )=P _dΔ ·(1-α)+(1-P _dΔ )·(1-β) (12)

P(F ₁ |H ₁ )=P _dΔ ·α+(1-P _dΔ )·β (13)

γ _P is the received signal-to-noise ratio of the main user at the cognitive user;

为PUEA对主用户进行能量检测时的检测概率：

Probability of detection when energy detection is performed for the primary user for PUEA:

Among them, γ _Δ is the received signal-to-noise ratio of the main user at the PUEA, and formulas (10) and (11) are substituted into formula (9) to obtain:

The false alarm error probability and missed detection error probability generated by each cognitive user in the detection channel and the reporting channel are:

P _fe =P _fc (1-P _e )+(1-P _fc )P _e (16)

P _me =P _m (1-P _e )+(1-P _m )P _e (17)

P_e为认知用户向融合中心发送判决结果过程中的传输差错概率。in,

P _e is the transmission error probability in the process of the cognitive user sending the judgment result to the fusion center.

In step C, use the K-rank optimization criterion for optimization, where K is the number of cognitive users required to complete the decision, and M is the number of all cognitive users in the cognitive network, then the global false alarm error probability P _F (K, M ) and the global missed detection error probability P _M (K, M) are:

Among them, P _r (D _{1 |} H _{0 )} is the probability that the main user does not exist but the fusion center judges that the main user exists _; probability of existence.

Find the global average error probability function of the system:

The derivation of the global average error probability of the cooperative spectrum sensing optimization system with respect to K can be obtained:

时可得：when

When available:

Taking the logarithm of both sides gives:

The K value can be obtained after calculation, and it is stipulated that the value obtained by rounding K backward is the number K ^* of cognitive users required by the system:

By substituting K ^* into the global average error probability of the system, the global average error probability of the system under the K-rank optimization criterion can be obtained.

Beneficial effects achieved by the present invention: Compared with several traditional schemes, the method of the present invention uses Cooperative Spectrum Sensing (CSS) to improve the sensing accuracy, including the detection channel and the reporting channel. Specifically, each CU makes a binary decision on the observed licensed frequency band through energy detection, and reports the decision result to the fusion center through the reporting channel. Finally, the fusion center makes a global decision through the K-rank optimization criterion: when at least When K CUs detect PU signals, the fusion center judges that the PU is occupying the licensed frequency band; if less than K CUs detect PU signals, the fusion center judges that the PU does not occupy the licensed frequency band. On the basis of this CSS model, the global mean error probability is minimized by optimizing the value of K. The present invention improves the accuracy of cognitive network spectrum sensing by optimizing the traditional fusion criterion.

Description of drawings

FIG. 1 is a model diagram of a cooperative spectrum sensing system based on cognitive radio technology according to an embodiment of the present invention;

Fig. 2 is the MATLAB simulation diagram of the global average error probability under different signal-to-noise ratios _γp of the embodiment in Fig. 1;

Figure 3 shows the comparison between the optimal K value solution obtained by the algorithm search and the K-rank optimization criterion under different signal-to-noise ratios _γp ;

Figure 4 shows the change of the K value in the K-rank optimization criterion under different signal-to-noise ratio _γp environments.

Detailed ways

The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

The technical means disclosed in the solution of the present invention are not limited to the technical means disclosed in the above embodiments, but also include technical solutions composed of any combination of the above technical features.

Example 1

The present invention provides a collaborative sensing optimization method for simulating primary user attacks on a cognitive network. The collaborative spectrum sensing optimization system includes a primary user (PU), a simulated primary user attacker (PUEA), and M cognitive users (CUs). and a data fusion center (FC), including the following steps:

In step A, the false alarm probability of the energy detection of a single cognitive user is used as a constraint condition to determine the judgment threshold of the energy detection of the cognitive user. probability;

Step B, derive the detection probability and missed detection probability formula of cognitive user energy detection, and obtain the global average error probability formula of the collaborative spectrum sensing optimization system, the detection probability exists for the main user and the detection result of the cognitive user is also: The probability that the main user exists, the missed detection probability is the probability that the main user exists and the detection result of the cognitive user does not exist;

Step C, using the K-rank optimization criterion to calculate the global average error probability;

The global average error probability of the cooperative spectrum sensing optimization system is the superposition of the detection channel error probability and the reporting channel error probability, where:

Detection channel error probability includes false alarm probability and missed detection probability;

The reported channel error probability includes the transmission error probability.

In step A, the following four situations correspond to the four channel states:

( _a ) S1 assumption: the primary user and PUEA do not exist, only the case of noise;

(b) S2 assumption: there are only primary users and noise _;

( _c ) S3 assumption: there are only PUEA users and noise;

( _d ) S4 assumption: the case where the primary user, the PUEA user, and the noise all exist.

The main user and cognitive user signals both obey the Gaussian distribution, and the noise signal is additive white Gaussian noise.

Assuming that the channel environment between all cognitive users and the primary user is the same, and the information received by the cognitive users through energy detection has the same signal-to-noise ratio, the above four channel states correspond respectively:

S ₁ ={F ₀ , H ₀ }

S ₂ ={F ₀ , H ₁ }

S ₃ ={F ₁ , H ₀ }

S ₄ ={F ₁ , H ₁ }#(1)

Among them, F ₀ indicates that PUEA does not exist, F ₁ indicates that PUEA exists, H ₀ indicates that the main user does not exist, and H ₁ indicates that the main user exists; if PUEA detects the existence of the main user, PUEA initiates a simulated main user attack with probability α. If PUEA does not detect the existence of the primary user, PUEA initiates a simulated primary user attack with probability β;

The false alarm probability P _fc represents the probability that the primary user does not exist but the detection result of the cognitive user shows that the primary user exists, that is, P _fc =P(D ₁ |H ₀ ), which can be obtained from the Bayesian formula:

Among them, D ₁ indicates that the main user exists in the result of cognitive user energy detection, D ₀ indicates that the main user does not exist in the result of cognitive user energy detection, and P(F ₀ |H ₀ ) indicates that the PUEA does not exist when the main user does not exist. The probability of existence, P(F ₁ |H ₀ ) represents the probability of the existence of PUEA when the primary user does not exist, and P(D ₁ |F ₀ , H ₀ ) does not exist as the primary user and PUEA, but the detection result of the cognitive user is the probability of the existence of the primary user, namely:

P(D ₁ |F ₁ ,H ₀ ) represents the probability that the primary user does not exist, the PUEA exists but the cognitive user detection result is the primary user, that is:

t为积分变量，λ_c为认知用户能量检测的判决阈值，γ_e为PUEA在认知用户处的接收信噪比；

表示背景噪声的方差；N为采样点数；其中，P(F₀|H₀)、P(F₁|H₀)的表达式分别为：in,

t is the integral variable, λ _c is the decision threshold of cognitive user energy detection, γ _e is the received signal-to-noise ratio of PUEA at the cognitive user;

represents the variance of background noise; N is the number of sampling points; among them, the expressions of P(F ₀ |H ₀ ) and P(F ₁ |H ₀ ) are:

P(F ₀ |H ₀ )=P _fΔ ·(1-α)+(1-P _fΔ )·(1-β) (5)

P(F ₁ |H ₀ )=P _fΔ ·α+(1-P _fΔ )·β (6)

为PUEA对主用户进行检测时的虚警概率：

False alarm probability when detecting primary users for PUEA:

λ _Δ is the decision threshold for the PUEA user to perform energy detection on the primary user.

In step B, substitute formulas (5) and (6) into formula (2) to obtain:

表示主用户存在、且认知用户的检测结果也显示主用户存在的概率，P(F₀|H₁)表示主用户存在时PUEA不存在的概率，P(F₁|H₁)表示主用户存在时PUEA也存在的概率，即

由贝叶斯公式可得：Similarly, the detection probability

represents the existence of the primary user, and the detection result of the cognitive user also shows the probability of the presence of the primary user, P(F ₀ |H ₁ ) represents the probability that the PUEA does not exist when the primary user exists, and P(F ₁ |H ₁ ) represents the primary user The probability that PUEA also exists when it exists, i.e.

It can be obtained from Bayesian formula:

P(D ₁ |F ₀ , H ₁ ) represents the probability that the primary user exists, the PUEA does not exist, and the cognitive user detection result exists as the primary user, namely:

P(D ₁ |F ₁ , H ₁ ) represents the probability that the primary user exists, the PUEA also exists, and the cognitive user detection result exists as the primary user, namely:

The expressions of P(F ₀ |H ₁ ) and P(F ₁ |H ₁ ) are:

P(F ₀ , H ₁ )=P _dΔ ·(1-α)+(1-P _dΔ )·(1-β) (12)

P(F ₁ ,H ₁ )=P _dΔ ·α+(1-P _dΔ )·β (13)

γ _P is the received signal-to-noise ratio of the main user at the cognitive user;

为PUEA对主用户进行能量检测时的检测概率：

Probability of detection when energy detection is performed for the primary user for PUEA:

Among them, γ _Δ is the received signal-to-noise ratio of the main user at the PUEA, and formulas (10) and (11) are substituted into formula (9) to obtain:

The false alarm error probability and missed detection error probability generated by each cognitive user in the detection channel and the reporting channel are:

P _fe =P _fc (1-P _e )+(1-P _fc )P _e (16)

P _me =P _m (1-P _e )+(1-P _m )P _e (17)

P_e为认知用户向融合中心发送判决结果过程中的传输差错概率。in,

P _e is the transmission error probability in the process of the cognitive user sending the judgment result to the fusion center.

In step C, use the K-rank optimization criterion for optimization, where K is the number of cognitive users required to complete the decision, and M is the number of all cognitive users in the cognitive network, then the global false alarm error probability P _F (K, M ) and the global missed detection error probability P _M (K, M) are:

Among them, P _r (D _{1 |} H _{0 )} is the probability that the main user does not exist but the fusion center judges that the main user exists _; probability of existence.

Find the global average error probability function of the system:

采样点数N＝10，约束条件为PUEA对主用户的虚警概塞

小于0.1，且认知用户对主用户的虚警概率

也小于0.1。令α＝0.2，β＝0.7，

P_e＝0.01，根据公式(7)，将

代入，可以求出PUEA对主用户进行能量检测的检测阈值λ_Δ，将λ_Δ代入公式(14)可以得到PUEA对主用户的检测概率

接着将

代入公式(8)可以求得认知用户对主用户进行能量检测的检测阈值λ_c；最后根据公式(15)可以求出认知用户对主用户的检测概率

In the process of MATLAB simulation in this embodiment, both γ _e and γ _Δ are 0dB, and the background noise

The number of sampling points is N=10, and the constraint condition is the false alarm of PUEA to the main user

Less than 0.1, and the false alarm probability of the cognitive user to the primary user

Also less than 0.1. Let α=0.2, β=0.7,

_Pe = 0.01, according to formula (7), the

Substitute in, the detection threshold λ _Δ of the energy detection of the primary user by PUEA can be obtained, and λ _Δ can be substituted into formula (14) to obtain the detection probability of the primary user by PUEA

Then will

Substitute into formula (8) to obtain the detection threshold λ _c of the cognitive user to the primary user for energy detection; finally, according to formula (15), the detection probability of the cognitive user to the primary user can be obtained

Figure 1 is a system model diagram.

As shown in Figure 2, given the false alarm probability, the global average error probability of the system decreases with the increase of the signal-to-noise ratio _γp . With the increase of signal-to-noise ratio _γp , the global average error probability of adopting or criterion tends to 0.34 after exceeding 0dB; the global average error probability of adopting and criterion tends to 0.1 after signal-to-noise ratio _γp exceeds 7dB; using majority criterion The global evaluation error probability tends to 1.1×10 ^-4 after the SNR _γp exceeds 5dB, and the global average error probability using the K-rank optimization criterion tends to 8.5×10 ^-6 after the SNR exceeds 7dB.

As shown in Figure 3, the algorithm search is compared with the K-rank optimization criterion, and the curves of the two completely coincide, which proves the effectiveness of the algorithm.

As shown in Figure 4, M=10, when the K-rank optimization criterion is adopted, the K value that minimizes the global average error probability of the system increases in a step-like manner with the increase of the signal-to-noise ratio. When the signal-to-noise ratio exceeds 5dB, K It is always equal to 7, which is consistent with Figure 1: after the signal-to-noise ratio γ _p exceeds 5dB, the value of K in the K-rank optimization criterion is equal to 7, while the value of K in most criteria is 6. Therefore, the K-rank optimization criterion is adopted. It is close to the global average error probability when the majority criterion is adopted.

Example 2

A collaborative sensing optimization method for simulating primary user attacks on cognitive networks. The collaborative spectrum sensing optimization system includes a primary user (PU), a simulated primary user attacker (PUEA), M cognitive users (CU) and a data Fusion Center (FC), including the following steps:

In step A, the false alarm probability of the energy detection of a single cognitive user is used as a constraint condition to determine the judgment threshold of the energy detection of the cognitive user. probability;

Step B, derive the detection probability and missed detection probability formula of cognitive user energy detection, and obtain the global average error probability formula of the collaborative spectrum sensing optimization system, the detection probability exists for the main user and the detection result of the cognitive user is also: The probability that the main user exists, the missed detection probability is the probability that the main user exists and the detection result of the cognitive user does not exist;

Step C, using the K-rank optimization criterion to calculate the global average error probability;

In step D, an iterative optimization algorithm is used to obtain the optimal K value to minimize the global average error probability.

The derivation of the global average error probability of the cooperative spectrum sensing optimization system with respect to K can be obtained:

时可得：when

When available:

Taking the logarithm of both sides gives:

The K value can be obtained after calculation, and it is stipulated that the value obtained by rounding K backward is the number K ^* of cognitive users required by the system:

By substituting K ^* into the global average error probability of the system, the global average error probability of the system under the K-rank optimization criterion can be obtained.

Other technical features are the same as in Embodiment 1.

Although embodiments of the present invention have been shown and described, it will be apparent to those skilled in the art that various changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principle and spirit of the invention , the scope of the invention is defined by the appended claims and their equivalents.

Claims (4)

1. A cooperative sensing optimization method for simulating master user attack for a cognitive network is characterized by comprising the following steps: the cooperative spectrum sensing optimization system comprises a master user (PU), a simulated master user attacker (PUEA), M Cognitive Users (CU) and a data Fusion Center (FC), and comprises the following steps:

step A, determining a judgment threshold value of the energy detection of a cognitive user by taking the false alarm probability of the energy detection of a single cognitive user as a constraint condition, wherein the false alarm probability is the probability that a main user does not exist and the detection result of the cognitive user is the probability that the main user exists;

step B, deducing a detection probability and a missed detection probability formula of the energy detection of the cognitive user, and solving a global average error probability formula of the cooperative spectrum sensing optimization system, wherein the detection probability is the probability that a main user exists and the detection result of the cognitive user also exists, and the missed detection probability is the probability that the main user exists and the detection result of the cognitive user does not exist;

step C, calculating the global average error probability by using a K rank optimization criterion;

in step C, optimizing by using a K rank optimization criterion, wherein K is the number of cognitive users required for finishing judgment, M is the number of all cognitive users in the cognitive network, and the global false alarm error probability P is obtained _F (K, M) and the global false miss probability P _M (K, M) are respectively:

wherein, P _r (D ₁ |H ₀ ) Probability P that the master user does not exist but the fusion center judgment result is the master user _r (D ₀ |H ₁ ) The probability that the master user exists but the fusion center judgment result is that the master user does not exist;

obtaining a system global average error probability function:

the global average error probability of the cooperative spectrum sensing optimization system is derived from K:

when in use

Then, the following can be obtained:

taking logarithm of two sides to obtain:

k value can be obtained through calculation, and the value obtained by rounding K backward is defined as the number K of the cognitive users required by the system ^* ：

Will K ^* Substituting the average error probability into the global average error probability of the system to obtain the global average error probability of the system under the K rank optimization criterion, P _fe 、P _me Respectively generating false alarm error probability and false missing detection error probability in a detection channel and a report channel for each cognitive user;

and D, obtaining the optimal K value by using an iterative optimization algorithm to minimize the global average error probability.

2. The cooperative sensing optimization method for simulating the attack of the master user by the cognitive network as claimed in claim 1, wherein the cooperative sensing optimization method comprises the following steps:

in step a, the following four cases correspond to four channel states, respectively:

(a)S ₁ suppose that: the master user and the PUEA do not exist, and only noise exists;

(b)S ₂ suppose that: only master users and noise exist;

(c)S ₃ suppose that: only the PUEA user and noise exist;

(d)S ₄ suppose that: the master user, the PUEA user and noise exist;

the signals of the master user and the cognitive user are subjected to Gaussian distribution, and the noise signal is additive white Gaussian noise;

assuming that channel environments between all cognitive users and a master user are the same, and information received by the cognitive users through energy detection has the same signal-to-noise ratio, the above four channel states respectively correspond to:

S ₁ ＝{F ₀ ，H ₀ }

S ₂ ＝{F ₀ ，H ₁ }

S ₃ ＝{F ₁ ，H ₀ }

S ₄ ＝{F ₁ ，H ₁ }#(1)

wherein, F ₀ Indicates the absence of PUEA, F ₁ Indicates the presence of PUEA, H ₀ Indicating the absence of primary user, H ₁ Indicating that a master user exists; if the PUEA detects that a master user exists, the PUEA initiates a simulated master user attack with a probability alpha, and if the PUEA does not detect that the master user exists, the PUEA initiates a simulated master user attack with a probability beta;

probability of false alarm P _fc The detection result indicating that the primary user does not exist but the cognitive user shows the probability that the primary user exists, i.e. P _fc ＝P(D ₁ |H ₀ ) From bayesian formula, we can obtain:

wherein D is ₁ Indicating the existence of a master user as a result of cognitive user energy detection, D ₀ The result of the cognitive user energy detection is that the master user does not exist, P (F) ₀ |H ₀ ) Indicating the probability that PUEA does not exist when a primary user does not exist, P (F) ₁ |H ₀ ) Indicating the probability of PUEA being present when a primary user is absent, P (D) ₁ |F ₀ ，H ₀ ) The probability that the main user and the PUEA do not exist but the detection result of the cognitive user is that the main user exists is as follows:

P(D ₁ |F ₁ ，H ₀ ) The probability that the main user does not exist, the PUEA exists and the cognitive user detection result is the main user exists is represented, namely:

wherein,

t is an integral variable, λ _c Decision threshold, gamma, for cognitive user energy detection _e The receiving signal-to-noise ratio of the PUEA at the cognitive user is obtained;

a variance representing background noise; n is the number of sampling points; wherein, P (F) ₀ |H ₀ )、P(F ₁ |H ₀ ) Are respectively:

P(F ₀ |H ₀ )＝P _fΔ ·(1-α)+(1-P _fΔ )·(1-β) (5)

P(F ₁ |H ₀ )＝P _fΔ ·α+(1-P _fΔ )·β (6)

false alarm probability when detecting the master user for PUEA:

λ _Δ and carrying out energy detection on the main user for the PUEA user.

3. The cooperative sensing optimization method for simulating the attack of the master user by the cognitive network as claimed in claim 2, wherein the cooperative sensing optimization method comprises the following steps: in step B, the global average error probability of the cooperative spectrum sensing optimization system is a superposition of the detection channel error probability and the reporting channel error probability, wherein:

the detection channel error probability comprises a false alarm probability and a missed detection probability;

reporting the channel error probability includes transmitting the error probability.

4. The cooperative sensing optimization method for simulating the attack of the master user by the cognitive network as claimed in claim 3, wherein the cooperative sensing optimization method comprises the following steps:

in step B, substituting equations (5) and (6) into equation (2) can obtain:

in the same way, the probability of detection

Probability that a primary user exists is shown as the detection result of the cognitive user and the primary user exists, P (F) ₀ |H ₁ ) Indicating the probability that PUEA does not exist when a primary user is present, P (F) ₁ |H ₁ ) Indicating the probability that a PUEA is present when a primary user is present,

the Bayesian formula can be used to obtain:

P(D ₁ |F ₀ ，H ₁ ) The probability that the main user exists, the PUEA does not exist and the cognitive user detection result is the main user exists is represented, namely:

wherein, γ _P For the received signal-to-noise ratio of the main user at the cognitive user, P (D) ₁ |F ₁ ，H ₁ ) The probability that the main user exists, the PUEA also exists and the cognitive user detection result is the main user exists is represented, namely:

P(F ₀ |H ₁ )、P(F ₁ |H ₁ ) Are respectively:

P(F ₀ |H ₁ )＝P _dΔ ·(1-α)+(1-P _dΔ )·(1-β) (12)

P(F ₁ |H ₁ )＝P _dΔ ·α+(1-P _dΔ )·β (13)

the detection probability when energy detection is carried out on the main user for PUEA is as follows:

wherein, γ _Δ Substituting equations (10), (11) into equation (9) for the received snr of the primary user at the PUEA can obtain:

the false alarm error probability and the false missing error probability generated by each cognitive user in the detection channel and the report channel are respectively as follows:

P _fe ＝P _fc (1-P _e )+(1-P _fc )P _e (16)

P _me ＝P _m (1-P _e )+(1-P _m )P _e (17)

wherein,

P _e and transmitting the transmission error probability in the process of sending the judgment result to the fusion center for the cognitive user.

Images